A LiDAR–Inertial SLAM Method Based on Virtual Inertial Navigation System

In scenarios with insufficient structural features, LiDAR-based SLAM may suffer from degeneracy, resulting in impaired robot localization and mapping potentially leading to subsequent deviant navigation tasks. Therefore, it is crucial develop advanced algorithms techniques mitigate the degeneracy issue ensure robustness accuracy of SLAM. This paper presents a LiDAR–inertial simultaneous (SLAM) method based on virtual inertial system (VINS) address degeneracy. We classified different gaits match each gait its corresponding torso measurement unit (IMU) sensor construct foot components. By combining an (INS) zero-velocity updates (ZUPTs), we formed VINS achieve real-time estimation correction. Finally, corrected pose was input IMU odometry calculation procedure further refine results. To evaluate effectiveness our proposed degenerate environments, conducted experiments three typical scenarios. The results demonstrate high suitability scenes show improvement point clouds effect. algorithm’s versatility emphasized by wide applicability GPU platforms, including quadruped robots human wearable devices. broader potential range applications extends other related fields such as autonomous driving.